Non-US institutions entering the United States settle structure early: a subsidiary, a partnership with a licensed party, an acquisition, or a referral arrangement. The choice is usually made on tax, funding or speed grounds.
It is also the choice that decides which regulatory environments the institution sits in, and therefore what every later product and state decision costs.
The same plan, four different institutions
Partnering puts consequence into a dependency and shortens time to market. Acquiring inherits a position, including the parts nobody described in the data room. Building standalone maximises control and front-loads the licensing work. Referral-only defers most of it and caps the business.
Each is legitimate. They are not interchangeable, and comparing them after the structure is in place is not a comparison at all — it is a justification.
Sequence follows structure
Once structure is fixed, state sequencing, product sequencing and partner selection all inherit its constraints. Testing structure first is what makes the rest tractable rather than reactive.
The interesting question for us is whether an institution can see all four versions of itself before committing to one, instead of discovering the other three later through cost.